TSMC’s A16 node brings backside power to its Angstrom era chips

TSMC is preparing to show more details about its A16 process at the 2026 VLSI Symposium, and the node looks like an important step beyond 2nm. A16 is part of TSMC’s Angstrom era roadmap and is expected to bring better performance, lower power use, and backside power delivery.

The biggest change is TSMC’s Super Power Rail technology. This is the company’s version of backside power delivery, which moves power routing to the back of the chip. That frees more space on the front side for signal routing and helps reduce power delivery losses.

A16 improves speed and power without changing everything from 2nm

A16 still uses nanosheet transistor technology, which TSMC is first introducing with its N2 process. The difference is that A16 adds Super Power Rail and further optimizations.

Compared with N2P, TSMC says A16 can offer an 8% to 10% speed improvement at the same voltage. It can also deliver a 15% to 20% power reduction at the same speed. Logic density and SRAM density are expected to improve by up to 8% to 10%.

A16 is aimed mainly at high performance computing chips, where dense power delivery and complex signal routing matter a lot. That includes AI accelerators, server CPUs, and other large chips where power efficiency and layout density can directly affect performance.

Mass production is planned for Q4 2026, but actual products using A16 are expected later, likely in the 2027 to 2028 period. That gap is normal because chip designers need time to finish designs, validate silicon, and move products through manufacturing.

TSMC is also lining up future nodes. A14 is expected to arrive after A16, while A13 and A12 are planned around 2029. A13 is described as a shrink of A14 with around 6% area savings and backward compatibility with A14. A12 is another A14 based improvement and is expected to use Super Power Rail as well.

The competitive angle is also important. Intel already uses backside power delivery with 18A, and TSMC is now preparing its own production path with A16. As AI demand keeps putting pressure on leading edge manufacturing capacity, both companies are trying to prove they can offer better performance, better efficiency, and enough capacity for large customers.

For normal PC buyers, this is still future technology. A16 chips will not appear overnight. But this is the kind of node that could shape future CPUs, GPUs, AI chips, and mobile processors. The main idea is clear: smaller transistors still matter, but better power delivery is now just as important.